The search for a single explanation for global mass extinctions has let to polarization and the controversies that are often fueled by widespread media attention. The historic record shows a roughly linear log-log relation between the frequency of explosive volcanic eruptions and the volume of their products. Eruptions such as Mt. St. Helens 1980 produce on the order of 1 cu km of tephra, destroying life over areas in the 10 to 100 sq km range, and take place, on the average, once or twice a decade. Eruptions producing 10 cu km take place several times a century and, like Krakatau 1883, destroy life over 100 to 1000 sq km areas while producing clear global atmospheric effects. Eruptions producting 10,000 cu km are known from the Quaternary record, and extrapolation from the historic record suggests that they occur perhaps once in 20,000 years, but none has occurred in historic time and little is known of their biologic effects. Even larger eruptions must also exist in the geologic record, but documentation of their volume becomes increasingly difficult as their age increases. The conclusion is inescapable that prehistoric eruptions have produced catastrophes on a global scale: only the magnitude of the associated mortality is in question. Differentiation of large magma chambers is on a time scale of thousands to millions of years, and explosive volcanoes are clearly concentrated in narrow belts near converging plate margins. Volcanism cannot be dismissed as a producer of global catastrophes. Its role in major extinctions is likely to be at least contributory and may well be large. More attention should be paid to global effects of the many huge eruptions in the geologic record that dwarf those known in historic time.